amp target selector dna egfr kit for tissue demonstrates ... · amplifiable dna in the tissue slide...

1
Target Selector DNA EGFR kit for ssue demonstrates high sensivity without the need for macro-dissecon Vassilios Alexiadis, PhD; Lyle J. Arnold, PhD; Jolly Shrivastava; Veena Singh, MD; Jason C. Poole, PhD Biocept Inc, San Diego CA AMP 2019 Category: Hematopathology Poster# #G021 Abstract# 16853 Introducon Technology Conclusions References Results Results Results Materials and Methods Biocept’s Target Selector ctDNA plaorm was originally developed to detect and track cancer-related mutaons (EGFR, KRAS, BRAF) from a simple blood draw for use in analyzing biomarkers from circulang tumor cells (CTCs) and circulang tumor DNA (ctDNA). We have previously reported extensive analycal performance characteriscs and clinical concordance of the Target Selector ctDNA assays. Here we demonstrate the expanded use of our Target Selector ctDNA assay on formalin- fixed paraffin-embedded ssue (FFPE). The high sensivity of Target Selector DNA assays (LOD, 0.01-0.03%) allows molecular analysis of pathological ssue secons without tumor content enrichment by macro-dissecon, potenally streamlining molecular tesng processes in a cost-effecve and me efficient manner. To test this kit, DNA was extracted from cancer paent FFPE ssue slides, analyzed by Target Selector qPCR assays, followed by Sanger sequencing. The mutaonal profile was compared to the pathological case report for each paent. FFPE DNA represents challenging material for molecular analyses of cancer-related mutaons because of DNA fragmentaon due to long-term storage as well as crosslinks (DNA:DNA and DNA:protein) due to formalin treatment that have to be reversed. Also, hydrolyc cytosine deaminaon which converts cytosine to uracil has been idenfied as a major source for arfacts in FFPE ssue. To overcome these challenges: First, we used EGFR exon 20 (WT) amplificaon as a normalizaon factor for amplifiable DNA in FFPE DNA. To this end, 3ul of DNA purified from non-macrodissected or macrodissected FFPE slides and from frozen normal ssue slides was performed in EGFR WT Target Selector . EGFR WT copy number was determined by performing on the same plate QPCR using a standard curve of serial diluons with EGFR copy numbers of 5000, 625, 78 and 10 respecvely. Second, the purificaon of FFPE DNA was done using an Uracil-N-Glycosylase treatment which eliminates uracil in U:G mismatches. Third, we established a cut-off for each Target Selector assay using the data from negave samples for each respecve mutaon Target Selector assay. Using the calculated cut-off of 1.7% for the T790M mutaon the T790M Target Selector assay has a sensivity of 100% (5/5 for both non-macrodissected and macrodissected) and a specificity of 96% (24/25) for non-macrodissected or 100% (25/25) for macro-dissected respecvely. The cutoff seng for each of the Target Selector assays are shown below in Fig 5 and the accuracy, sensivity and specificity for each assay based on these cutoffs is summarized in Fig 6. DNA from 30 cancer paent cases (prepared from either non- macrodissected or macrodissected FFPE slides) or from 3 normal ssue slide was then used in Target Selector assays specific for mutaons in Biomarkers: T790M, L858R, Del19, BRAF V600 or KRAS codon 12/13. A standard curve was used for each mutant assay to quanfy the mutant copies and the rao of mutant copies to EGFR WT copies was used to determine the mutant allele frequency. The data from negave samples for each mutant assay were tabulated to establish a cut-off and determine sensivity and specificity for each assay using the cut-off value. The cut-off was set as the mean of the allele frequency (%) + 3*Standard deviaons. All QPCR reacons showing amplificaon by Target Selector were analyzed by Sanger sequencing for confirmaon of presence of mutant. Analysis of Target Selector mutaon data obtained using FFPE DNA from non-macrodissected or macro-dissected ssue slides indicate that specific mutaons for paent cases were detected with high sensivity and specificity. Because of arfacts introduced by formaldehyde fixaon and removal for FFPE DNA preparaon, cutoffs were established for each of the mutant Target Selector assays based on FFPE ssue not containing the mutaon or normal frozen ssue slides. FFPE DNA is a suitable input for Target Selector assays and allows mutaon analysis without prior macro-dissecon of FFPE tumor secons. 1. Poole JC, et al. (2019) Analycal validaon of the Target Selector ctDNA plaorm featuring single copy detecon sensivity for clinically aconable EGFR, BRAF, and KRAS mutaons. PLoS ONE 14(10): e0223112. hps://doi. org/10.1371/journal.pone.0223112 2. Kim, S. S., et al. (2018) Droplet digital PCR-based EGFR mutaon detecon with an internal quality control index to determine the quality of DNA. Scienfic reports, 8(1), 543. doi:10.1038/s41598-017-18642-x Unstained FFPE slides (5µm secons) and corresponding H&E slide from tumor cases with EGFR (L858R, Del19 or T790M), BRAF V600 or KRAS mutaons (codon 12/13) were obtained from Discovery Life Sciences or Reprocell. The H&E slides were analyzed by in-house pathologist and marked for tumor content. The tumor-marked H&E slides were used to mark unstained slides for each case for macrodissecon. Frozen slides from normal ssue 5µm secons (lung, colon or breast obtained from Origene) were also used as negave controls. DNA was extracted from the FFPE or normal ssue using QS GeneRead DNA FFPE treatment and QIAsymphony DSP DNA Mini kits on a QIAsymphony instrument (QIAGEN). Purified DNA was used in Target Selector reacons on QS5 followed by Sanger sequencing (3500Dx) to confirm mutaons. EGFR exon 20 (WT) amplificaon was also performed in addion to mutant tesng for normalizaon of the data to amplifiable DNA in the ssue slide samples. Figure 1. FFPE ssue from unstained slides was scraped for DNA extracon either from whole slide (non-macrodissected) or only from area marked by pathologist as tumor on corresponding H&E slide (macro-dissected). Figure 2. Summary of cancer paent cases analyzed with Target Selector assays using FFPE ssue slides. Clinical characteriscs, tumor content and number of cases for each of the mutant EGFR L858R, del19, T790M, BRAF V600 and KRAS codon 12/13 Biomarker paent cases are indicated. Figure 3: A) Target Selector kit assays ulize qPCR followed by DNA sequencing to verify mutaons. B) For FFPE Target Selector kit assays FFPE DNA is prepared on QIAsymphony and used in Target Selector mutant enrichment. C) Target Selector mutant enrichment applies a blocker (switch + anchor) to block WT DNA amplificaon while allowing mutant DNA amplificaon. Figure 6: Target Selector cut-off seng is indicated for each Biomarker and was used to determine accuracy, sensivity and specificity of Target Selector assays using FFPE ssue (non-macrodissected or macrodissected). A. Target Selector kit workflow Target Selector cut-off seng for FFPE ssue C. Target Selector™ mutant enrichment Target Selector™ assays cut-off seng FFPE DNA Summary: Accuracy, Sensivity and Specificity Biomarker Cut-off (%) Accuracy Sensivity Specificity Non- macrodissected Macrodissected Non- macrodissected Macrodissected Non- macrodissected Macrodissected L858R 0.17 97% (29/30) 93% (28/30) 100% (16/16) 94% (15/16) 93% (13/14) 93% (13/14) Del19 0.17 100% (30/30) 100% (30/30) 100% (5/5) 100% (5/5) 100% (25/25) 100% (25/25) T790M 1.7 97% (29/30) 100% (30/30) 100% (5/5) 100% (5/5) 96% (24/25) 100% (25/25) BRAF 0.46 100% (30/30) 100% (30/30) 100% (5/5) 100% (5/5) 100% (25/25) 100% (25/25) KRAS 1.64 93% (28/30) 100% (30/30) 100% (4/4) 100% (4/4) 92% (24/26) 100% (26/26) Total N/A 97.3% (146/150) 98.7% (148/150) 100% (35/35) 97.1% (34/35) 96.5% (111/115) 99.1% (114/115) Figure 5. Target Selector cut-offs using negave cases (FFPE slides and frozen ssue) for T790M, L858R, Del19, KRAS and BRAF were set at 1.71%, 0.16%, 0.17%, 1.64% and 0.46% respecvely (TP: True Posive; FP: False Posive; TN: True Negave; FN: False Negave). [Note: Del#3 has been confirmed as L858R case by vendor and confirmed in this study by Sanger sequencing.] Target Selector™ assay ssue results Sanger sequencing summary Biomarker Paent case Macro- dissected Sanger Sequencing results T790M L858R Del19 BRAF KRAS Del19 1 - T790M L858R Del 747-753 ins Ser V600 syn. mut. N/A + T790M L858 and G857 syn. mut. Del 747-753 ins Ser N/A N/A 2 - T790M L858 and G857 syn. mut. Del746-750 N/A N/A + T790M L858 and G857 syn. mut. Del746-750 V600 syn. mut. N/A 3 - T790M L858R N/A Not readable G12D + T790M L858R N/A V600 syn. mut. N/A 4 - T790M L858R, G857 syn. mut. Del 746-750 V600 syn. mut. N/A + T790M L858 and G857 syn. mut. Del 746-750 Not readable N/A 5 - T790M G857 syn mut. Del 747-752 ins Ser V600 syn. mut. Not readable + T790M N/A Del 747-752 ins Ser K601 syn. mut. Not readable 6 - T790M G857V Del747-753 ins Ser N/A G12A + T790 syn. mut. N/A Del747-753 ins Ser N/A N/A L858R 1 - T790M L858R N/A V600L Not readable + T790M L858R N/A V600 syn. mut. Not readable 2 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A 3 - T790M L858R N/A Not readable N/A + T790M L858R N/A Not readable N/A 4 - T790M L858R N/A Not readable N/A + T790M L858R Del 746-750 N/A N/A 5 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A 6 - T790M L858R N/A V600 syn. mut. N/A + T790M L858R N/A N/A N/A 7 - T790M L858R N/A V600M N/A + T790M L858R N/A V600L Not readable 8 - T790M L858R N/A N/A N/A + N/A N/A N/A N/A N/A 9 - T790M L858R N/A V600M N/A + T790M L858R N/A N/A N/A 10 - T790M L858R N/A V600E N/A + T790M L858R N/A N/A N/A T790M 1 - T790M L858R N/A S602Y G13S + T790M L858R N/A V600L N/A 2 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A 3 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A 4 - T790M L858R Del746-Del750 S602Y G12D + T790M L858R N/A N/A Not readable 5 - T790M L858R N/A V600 syn. mut. N/A + T790M L858R N/A N/A N/A BRAF 1 - T790M L858 syn. mut. N/A V600E N/A + T790M L858 syn. mut. N/A V600E Not readable 2 - T790M L858 syn. mut. N/A V600E N/A + T790M N/A N/A V600E N/A 3 - T790M L858 syn. mut. and G857V N/A V600E N/A + N/A N/A N/A V600E N/A 4 - T790M L858 syn. mut and L858F N/A V600E N/A + T790M L858 syn. mut N/A V600E Not readable 5 - T790M L858 and G857 syn. mut. N/A V600E N/A + T790M L858 syn. mut. N/A V600E N/A KRAS 1 - T790M L858R N/A N/A G12D + T790M N/A N/A N/A G12D 2 - T790M L858R, L858 syn. mut. N/A N/A G13D + T790M L858R N/A N/A G13D 3 - T790M N/A N/A K601 syn. mut. G12V + T790M L858R, L858 syn. mut, G857 syn. mut. N/A S602F G12V 4 - T790M L858 syn. mut. N/A N/A G12D + T790M N/A N/A N/A G12D Normal 1 Normal colon T790M L858 syn. mut. N/A V600M N/A 2 Normal lung T790M L858 syn. mut. N/A V600C ins Gly Not readable 3 Normal breast T790M L858 syn. mut. N/A V600M N/A Clinical characteriscs of cancer paent cases Sample number Age, paent (years) Gender Tumor type Mutaon Clinical stage Age, FFPE slides (years) 1 69 F NSCLC L858R I-A 20 2 79 M NSCLC L858R III-A 14 3 79 M NSCLC L858R III-A 14 4 60 F NSCLC L858R Primary 12 5 79 M NSCLC L858R III-A 14 6 58 M NSCLC L858R I-B 6 7 53 F NSCLC L858R I-A 4 8 59 M NSCLC L858R II-A 6 9 72 F NSCLC L858R I-B 5 10 52 F NSCLC L858R I-B 5 11 56 M NSCLC L858R I-A 5 12 36 M NSCLC T790M/L858R I-B 1 13 49 M NSCLC T790M/L858R I-A 1 14 51 M NSCLC T790M/L858R I-A 1 15 45 M NSCLC T790M/L858R I-A 3 16 67 F NSCLC T790M/L858R I-B 2 17 68 F NSCLC Del19 I-A 16 18 48 F NSCLC Del19 I-A 7 19 55 M NSCLC Del19 II-A 5 20 68 F NSCLC Del19 I-A 6 21 61 F NSCLC Del19 I-B 6 22 Unknown F CRC BRAF II-A 23 23 Unknown Unknown CRC BRAF Unknown 23 24 42 M CRC BRAF Unknown 24 25 Unknown Unknown CRC BRAF Unknown 24 26 57 M Melanoma BRAF I-B 21 27 78 F CRC KRAS Unknown 17 28 58 F CRC KRAS Unknown 18 29 78 F NSCLC KRAS II-A 15 30 81 F CRC KRAS III-C 19 Tumor content of FFPE slides Number of cases tested for each Biomarker L858R Del19 T790M BRAF KRAS 100 80 60 40 20 0 Tumor content (%) Sample number 1 5 10 15 20 25 30 0 2 4 6 8 10 12 14 16 L858R T790M Del19 BRAF KRAS None Number of cases tested Biomarker Frozen normal Melanoma CRC NSCLC B. Target Selector FFPE slide experimental procedure Sanger sequencing (3500Dx) Target Selector™ qPCR product with appropriate sequencing primer Target Selector™ EGFR kit (qPCR on QS5) EGFR WT T790M L858R Del19 KRAS 12/13 BRAF V600E DNA preparaon from FFPE slides No Macrodissecon Macrodissecon Figure 4. (A) Standard curve which was used to quanfy EGFR WT copies in FFPE DNA or normal ssue DNA. Serial diluon of genomic DNA from H1975 cell line corresponding to 5000, 625, 78 or 10 copies of EGFR respecvely was used in EGFR WT Target Selector . The standard curve was used to quanfy the EGFR copies in DNA extracted from FFPE DNA or normal ssue DNA which were run on the same QPCR plate. (B) EGFR copies determined by Target Selector for FFPE DNA (from non-macro-dissected or macro-dissected ssue) or normal ssue. The total copies for each paent case are indicated. The total copies EGFR (50ul) were used in normalizaon of mutant data and analysis of mutant allele frequency for each of Target Selector assays tested. Paent case Total copies EGFR (50ul) Non-macro Macro Del19 #1 12688 3636 Del19 #2 51875 65986 Del19 #3 7343 3561 Del19 #4 19339 21915 Del19 #5 5151 4182 Del19 #6 10789 1946 L858R#1 7335 5965 L858R#2 637 353 L858R#3 34479 13712 L858R#4 22033 8446 L858R#5 1999 1846 L858R#6 51853 27294 L858R#7 49773 37203 L858R#8 2439 0 L858R#9 4453 223 L858R#10 10408 1513 T790M#1 7341 4561 T790M#2 3171 2900 T790M#3 4171 586 T790M#4 7044 3670 T790M#5 16052 6435 BRAF #1 15225 8593 BRAF #2 2863 935 BRAF #3 3804 1297 BRAF #4 17223 8733 BRAF #5 10957 3963 KRAS #1 4035 2378 KRAS #2 4032 2725 KRAS #3 8481 4302 KRAS #4 1626 2129 Normal colon 43336 N/A Normal lung 37137 N/A Normal breast 78776 N/A A B True posive False posive True negave False negave N/A: Not applicable (No qPCR amplificaon detected) syn. mut.=synonymous mutaon

Upload: others

Post on 22-Sep-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: AMP Target Selector DNA EGFR kit for tissue demonstrates ... · amplifiable DNA in the tissue slide samples. Figure 1. FFPE tissue from unstained slides was scraped for DNA extraction

Target Selector™ DNA EGFR kit for tissue demonstrates high sensitivity without the need for macro-dissection

Vassilios Alexiadis, PhD; Lyle J. Arnold, PhD; Jolly Shrivastava; Veena Singh, MD; Jason C. Poole, PhD

Biocept Inc, San Diego CA

AMP

2019

Category: Hematopathology

Poster# #G021

Abstract# 16853

Introduction Technology

Conclusions

References

Results Results Results

Materials and Methods

Biocept’s Target Selector™ ctDNA platform was originally developed to detect and track cancer-related mutations (EGFR, KRAS, BRAF) from a simple blood draw for use in analyzing biomarkers from circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA). We have previously reported extensive analytical performance characteristics and clinical concordance of the Target Selector™ ctDNA assays. Here we demonstrate the expanded use of our Target Selector™ ctDNA assay on formalin-fixed paraffin-embedded tissue (FFPE). The high sensitivity of Target Selector™ DNA assays (LOD, 0.01-0.03%) allows molecular analysis of pathological tissue sections without tumor content enrichment by macro-dissection, potentially streamlining molecular testing processes in a cost-effective and time efficient manner. To test this kit, DNA was extracted from cancer patient FFPE tissue slides, analyzed by Target Selector™ qPCR assays, followed by Sanger sequencing. The mutational profile was compared to the pathological case report for each patient.

FFPE DNA represents challenging material for molecular analyses of cancer-related mutations because of DNA fragmentation due to long-term storage as well as crosslinks (DNA:DNA and DNA:protein) due to formalin treatment that have to be reversed. Also, hydrolytic cytosine deamination which converts cytosine to uracil has been identified as a major source for artifacts in FFPE tissue. To overcome these challenges: First, we used EGFR exon 20 (WT) amplification as a normalization factor for amplifiable DNA in FFPE DNA. To this end, 3ul of DNA purified from non-macrodissected or macrodissected FFPE slides and from frozen normal tissue slides was performed in EGFR WT Target Selector™. EGFR WT copy number was determined by performing on the same plate QPCR using a standard curve of serial dilutions with EGFR copy numbers of 5000, 625, 78 and 10 respectively. Second, the purification of FFPE DNA was done using an Uracil-N-Glycosylase treatment which eliminates uracil in U:G mismatches. Third, we established a cut-off for each Target Selector™ assay using the data from negative samples for each respective mutation Target Selector™ assay.

Using the calculated cut-off of 1.7% for the T790M mutation the T790M Target Selector™ assay has a sensitivity of 100% (5/5 for both non-macrodissected and macrodissected) and a specificity of 96% (24/25) for non-macrodissected or 100% (25/25) for macro-dissected respectively. The cutoff setting for each of the Target Selector assays are shown below in Fig 5 and the accuracy, sensitivity and specificity for each assay based on these cutoffs is summarized in Fig 6.

DNA from 30 cancer patient cases (prepared from either non-macrodissected or macrodissected FFPE slides) or from 3 normal tissue slide was then used in Target Selector assays specific for mutations in Biomarkers: T790M, L858R, Del19, BRAF V600 or KRAS codon 12/13. A standard curve was used for each mutant assay to quantify the mutant copies and the ratio of mutant copies to EGFR WT copies was used to determine the mutant allele frequency. The data from negative samples for each mutant assay were tabulated to establish a cut-off and determine sensitivity and specificity for each assay using the cut-off value. The cut-off was set as the mean of the allele frequency (%) + 3*Standard deviations. All QPCR reactions showing amplification by Target Selector™ were analyzed by Sanger sequencing for confirmation of presence of mutant.

• Analysis of Target Selector™ mutation data obtained using FFPE DNA from non-macrodissected or macro-dissected tissue slides indicate that specific mutations for patient cases were detected with high sensitivity and specificity.

• Because of artifacts introduced by formaldehyde fixation and removal for FFPE DNA preparation, cutoffs were established for each of the mutant Target Selector™ assays based on FFPE tissue not containing the mutation or normal frozen tissue slides.

• FFPE DNA is a suitable input for Target Selector™ assays and allows mutation analysis without prior macro-dissection of FFPE tumor sections.

1. Poole JC, et al. (2019) Analytical validation of the Target Selector ctDNA platform featuring single copy detection sensitivity for clinically actionable EGFR, BRAF, and KRAS mutations. PLoS ONE 14(10): e0223112. https://doi.org/10.1371/journal.pone.0223112

2. Kim, S. S., et al. (2018) Droplet digital PCR-based EGFR mutation detection with an internal quality control index to determine the quality of DNA. Scientific reports, 8(1), 543. doi:10.1038/s41598-017-18642-x

Unstained FFPE slides (5µm sections) and corresponding H&E slide from tumor cases with EGFR (L858R, Del19 or T790M), BRAF V600 or KRAS mutations (codon 12/13) were obtained from Discovery Life Sciences or Reprocell. The H&E slides were analyzed by in-house pathologist and marked for tumor content. The tumor-marked H&E slides were used to mark unstained slides for each case for macrodissection.

Frozen slides from normal tissue 5µm sections (lung, colon or breast obtained from Origene) were also used as negative controls. DNA was extracted from the FFPE or normal tissue using QS GeneRead DNA FFPE treatment and QIAsymphony DSP DNA Mini kits on a QIAsymphony instrument (QIAGEN). Purified DNA was used in Target Selector™ reactions on QS5 followed by Sanger sequencing (3500Dx) to confirm mutations. EGFR exon 20 (WT) amplification was also performed in addition to mutant testing for normalization of the data to amplifiable DNA in the tissue slide samples.

Figure 1. FFPE tissue from unstained slides was scraped for DNA extraction either from whole slide (non-macrodissected) or only from area marked by pathologist as tumor on corresponding H&E slide (macro-dissected).

Figure 2. Summary of cancer patient cases analyzed with Target Selector™ assays using FFPE tissue slides. Clinical characteristics, tumor content and number of cases for each of the mutant EGFR L858R, del19, T790M, BRAF V600 and KRAS codon 12/13 Biomarker patient cases are indicated.

Figure 3: A) Target Selector™ kit assays utilize qPCR followed by DNA sequencing to verify mutations. B) For FFPE Target Selector™ kit assays FFPE DNA is prepared on QIAsymphony and used in Target Selector™ mutant enrichment. C) Target Selector™ mutant enrichment applies a blocker (switch + anchor) to block WT DNA amplification while allowing mutant DNA amplification.

Figure 6: Target Selector™ cut-off setting is indicated for each Biomarker and was used to determine accuracy, sensitivity and specificity of Target Selector™ assays using FFPE tissue (non-macrodissected or macrodissected).

A. Target Selector™ kit workflow Target Selector™ cut-off setting for FFPE tissue

C. Target Selector™ mutant enrichment

Target Selector™ assays cut-off setting

FFPE DNA Summary: Accuracy, Sensitivity and Specificity

Biomarker Cut-off (%)

Accuracy Sensitivity Specificity

Non-macrodissected Macrodissected Non-

macrodissected Macrodissected Non-macrodissected Macrodissected

L858R 0.17 97% (29/30) 93% (28/30) 100% (16/16) 94% (15/16) 93% (13/14) 93% (13/14)

Del19 0.17 100% (30/30) 100% (30/30) 100% (5/5) 100% (5/5) 100% (25/25) 100% (25/25)

T790M 1.7 97% (29/30) 100% (30/30) 100% (5/5) 100% (5/5) 96% (24/25) 100% (25/25)

BRAF 0.46 100% (30/30) 100% (30/30) 100% (5/5) 100% (5/5) 100% (25/25) 100% (25/25)

KRAS 1.64 93% (28/30) 100% (30/30) 100% (4/4) 100% (4/4) 92% (24/26) 100% (26/26)

Total N/A 97.3% (146/150) 98.7% (148/150) 100% (35/35) 97.1% (34/35) 96.5% (111/115) 99.1% (114/115)

Figure 5. Target Selector™ cut-offs using negative cases (FFPE slides and frozen tissue) for T790M, L858R, Del19, KRAS and BRAF were set at 1.71%, 0.16%, 0.17%, 1.64% and 0.46% respectively (TP: True Positive; FP: False Positive; TN: True Negative; FN: False Negative). [Note: Del#3 has been confirmed as L858R case by vendor and confirmed in this study by Sanger sequencing.]

Target Selector™ assay tissue results Sanger sequencing summary

BiomarkerPatient

caseMacro-

dissectedSanger Sequencing results

T790M L858R Del19 BRAF KRAS

Del19

1 - T790M L858R Del 747-753 ins Ser V600 syn. mut. N/A + T790M L858 and G857 syn. mut. Del 747-753 ins Ser N/A N/A

2 - T790M L858 and G857 syn. mut. Del746-750 N/A N/A + T790M L858 and G857 syn. mut. Del746-750 V600 syn. mut. N/A

3 - T790M L858R N/A Not readable G12D + T790M L858R N/A V600 syn. mut. N/A

4 - T790M L858R, G857 syn. mut. Del 746-750 V600 syn. mut. N/A + T790M L858 and G857 syn. mut. Del 746-750 Not readable N/A

5 - T790M G857 syn mut. Del 747-752 ins Ser V600 syn. mut. Not readable + T790M N/A Del 747-752 ins Ser K601 syn. mut. Not readable

6 - T790M G857V Del747-753 ins Ser N/A G12A + T790 syn. mut. N/A Del747-753 ins Ser N/A N/A

L858R

1 - T790M L858R N/A V600L Not readable + T790M L858R N/A V600 syn. mut. Not readable

2 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A

3 - T790M L858R N/A Not readable N/A + T790M L858R N/A Not readable N/A

4 - T790M L858R N/A Not readable N/A + T790M L858R Del 746-750 N/A N/A

5 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A

6 - T790M L858R N/A V600 syn. mut. N/A + T790M L858R N/A N/A N/A

7 - T790M L858R N/A V600M N/A + T790M L858R N/A V600L Not readable

8 - T790M L858R N/A N/A N/A + N/A N/A N/A N/A N/A

9 - T790M L858R N/A V600M N/A + T790M L858R N/A N/A N/A

10 - T790M L858R N/A V600E N/A + T790M L858R N/A N/A N/A

T790M

1 - T790M L858R N/A S602Y G13S + T790M L858R N/A V600L N/A

2 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A

3 - T790M L858R N/A N/A N/A + T790M L858R N/A N/A N/A

4 - T790M L858R Del746-Del750 S602Y G12D + T790M L858R N/A N/A Not readable

5 - T790M L858R N/A V600 syn. mut. N/A + T790M L858R N/A N/A N/A

BRAF

1 - T790M L858 syn. mut. N/A V600E N/A + T790M L858 syn. mut. N/A V600E Not readable

2 - T790M L858 syn. mut. N/A V600E N/A + T790M N/A N/A V600E N/A

3 - T790M L858 syn. mut. and G857V N/A V600E N/A + N/A N/A N/A V600E N/A

4 - T790M L858 syn. mut and L858F N/A V600E N/A + T790M L858 syn. mut N/A V600E Not readable

5 - T790M L858 and G857 syn. mut. N/A V600E N/A + T790M L858 syn. mut. N/A V600E N/A

KRAS

1 - T790M L858R N/A N/A G12D + T790M N/A N/A N/A G12D

2 - T790M L858R, L858 syn. mut. N/A N/A G13D + T790M L858R N/A N/A G13D

3 - T790M N/A N/A K601 syn. mut. G12V + T790M L858R, L858 syn. mut,

G857 syn. mut. N/A S602F G12V

4 - T790M L858 syn. mut. N/A N/A G12D + T790M N/A N/A N/A G12D

Normal

1 Normal colon T790M L858 syn. mut. N/A V600M N/A

2 Normal lung T790M L858 syn. mut. N/A V600C ins Gly Not readable

3 Normal breast T790M L858 syn. mut. N/A V600M N/A

Clinical characteristics of cancer patient cases Sample number

Age, patient (years) Gender Tumor type Mutation Clinical stage Age, FFPE

slides (years)

1 69 F NSCLC L858R I-A 202 79 M NSCLC L858R III-A 143 79 M NSCLC L858R III-A 144 60 F NSCLC L858R Primary 125 79 M NSCLC L858R III-A 146 58 M NSCLC L858R I-B 67 53 F NSCLC L858R I-A 48 59 M NSCLC L858R II-A 69 72 F NSCLC L858R I-B 5

10 52 F NSCLC L858R I-B 511 56 M NSCLC L858R I-A 512 36 M NSCLC T790M/L858R I-B 113 49 M NSCLC T790M/L858R I-A 114 51 M NSCLC T790M/L858R I-A 115 45 M NSCLC T790M/L858R I-A 316 67 F NSCLC T790M/L858R I-B 217 68 F NSCLC Del19 I-A 1618 48 F NSCLC Del19 I-A 719 55 M NSCLC Del19 II-A 520 68 F NSCLC Del19 I-A 621 61 F NSCLC Del19 I-B 622 Unknown F CRC BRAF II-A 2323 Unknown Unknown CRC BRAF Unknown 2324 42 M CRC BRAF Unknown 2425 Unknown Unknown CRC BRAF Unknown 2426 57 M Melanoma BRAF I-B 2127 78 F CRC KRAS Unknown 1728 58 F CRC KRAS Unknown 1829 78 F NSCLC KRAS II-A 1530 81 F CRC KRAS III-C 19

Tumor content of FFPE slides

Number of cases tested for each Biomarker

L858R Del19T790M BRAF KRAS100

80

60

40

20

0

Tum

or c

onte

nt (%

)

Sample number1 5 10 15 20 25 30

0

2

46

8

1012

14

16

L858R T790M Del19 BRAF KRAS None

Num

ber o

f cas

es te

sted

Biomarker

Frozen normalMelanomaCRCNSCLC

B. Target Selector™ FFPE slide experimental procedure

Sanger sequencing (3500Dx) Target Selector™ qPCR product with appropriate sequencing primer

Target Selector™ EGFR kit (qPCR on QS5) EGFR WT T790M L858R Del19 KRAS 12/13 BRAF V600E

DNA preparation from FFPE slides

No Macrodissection Macrodissection

Figure 4. (A) Standard curve which was used to quantify EGFR WT copies in FFPE DNA or normal tissue DNA. Serial dilution of genomic DNA from H1975 cell line corresponding to 5000, 625, 78 or 10 copies of EGFR respectively was used in EGFR WT Target Selector™. The standard curve was used to quantify the EGFR copies in DNA extracted from FFPE DNA or normal tissue DNA which were run on the same QPCR plate. (B) EGFR copies determined by Target Selector™ for FFPE DNA (from non-macro-dissected or macro-dissected tissue) or normal tissue. The total copies for each patient case are indicated. The total copies EGFR (50ul) were used in normalization of mutant data and analysis of mutant allele frequency for each of Target Selector™ assays tested.

Patient caseTotal copies EGFR (50ul)

Non-macro Macro

Del19 #1 12688 3636Del19 #2 51875 65986Del19 #3 7343 3561Del19 #4 19339 21915Del19 #5 5151 4182Del19 #6 10789 1946L858R#1 7335 5965L858R#2 637 353L858R#3 34479 13712L858R#4 22033 8446L858R#5 1999 1846L858R#6 51853 27294L858R#7 49773 37203L858R#8 2439 0L858R#9 4453 223

L858R#10 10408 1513T790M#1 7341 4561T790M#2 3171 2900T790M#3 4171 586T790M#4 7044 3670T790M#5 16052 6435BRAF #1 15225 8593BRAF #2 2863 935BRAF #3 3804 1297BRAF #4 17223 8733BRAF #5 10957 3963KRAS #1 4035 2378KRAS #2 4032 2725KRAS #3 8481 4302KRAS #4 1626 2129

Normal colon 43336 N/ANormal lung 37137 N/A

Normal breast 78776 N/A

A B

True positive

False positive

True negative

False negative

N/A: Not applicable (No qPCR amplification detected)syn. mut.=synonymous mutation